Ready 5.6.6 (#2270)

* Update README.md

* Bump version from 5.6.4 to 5.6.6

Author: Qubitium

README.md

@@ -17,6 +17,7 @@
 </p>
 
 ## Latest News
+* 12/15/2025 [5.6.6 Patch](https://github.com/ModelCloud/GPTQModel/releases/tag/v5.6.6): Fixed `multi-arch` build issues and `Tritonv2` kernel launch bug on multi-gpus. 
 * 12/12/2025 [5.6.2 Patch](https://github.com/ModelCloud/GPTQModel/releases/tag/v5.6.2): Fixed 3-bit Triton GPTQ kernel dequant/inference and `license` property compatibility issue with latest pip/setuptools.
 * 12/9/2025 [5.6.0](https://github.com/ModelCloud/GPTQModel/releases/tag/v5.6.0): ✨New `HF Kernel` for CPU optimized for `AMX`, `AVX2` and `AVX512`. Auto module tree for auto-model support. Added Afmoe and Dosts1 model support. Fixed pre-layer pass quantization speed regression. Improved HF Transformers, Peft and Optimum support for both GPTQ and AWQ. Fixed many AWQ compat bugs and regressions. 
 * 11/9/2025 [5.4.0](https://github.com/ModelCloud/GPTQModel/releases/tag/v5.4.0): ✨New Intel CPU and XPU hw optimized AWQ `TorchFusedAWQ` kernel. Torch Fused kernels now compatible with `torch.compile`. Fixed AWQ MoE model compatibility and reduced vram usage.

gptqmodel/version.py

@@ -7,4 +7,4 @@
 # even minor versions are release
 # 5.2.0 => release, 5.1.0 => devel
 # micro version (5.2.x) denotes patch fix, i.e. 5.2.1 is a patch fix release
-__version__ = "5.6.4"
+__version__ = "5.6.6"